Empirical Orthogonal Function Analysis and Modeling of Global Tropospheric Delay Spherical Harmonic Coefficients

Abstract

Based on the ERA-5 meteorological data from 2015 to 2019, we establish the global tropospheric delay spherical harmonic (SH) coefficients set called the SH_set and develop the global tropospheric delay SH coefficients empirical model called EGtrop using the empirical orthogonal function (EOF) method and periodic functions. We apply tropospheric delay derived from IGS stations not involved in modeling as reference data for validating the dataset, and statistical results indicate that the global mean Bias of the SH_set is 0.08 cm, while the average global root mean square error (RMSE) is 2.61 cm, which meets the requirements of the tropospheric delay model applied in the wide-area augmentation system (WAAS), indicating the feasibility of the product strategy. The tropospheric delay calculated with global sounding station and tropospheric delay products of IGS stations in 2020 are employed to validate the new product model. It is verified that the EGtrop model has high accuracy with Bias and RMSE of −0.25 cm and 3.79 cm, respectively, with respect to the sounding station, and with Bias and RMSE of 0.42 cm and 3.65 cm, respectively, with respect to IGS products. The EGtrop model is applicable not only at the global scale but also at the regional scale and exhibits the advantage of local enhancement.

Determination of the Availability and Variation of Surface Downwelling Shortwave Radiation in Saudi Arabia Using EUMETSAT Satellite Imagery

“…In other words, possible spatial modes of variability (patterns of variability) in climate studies can often be studied by EOF analysis. It is used for decomposing the time series data set into mutually independent spatial and temporal function parts to explain the spatiotemporal variability with the least possible modes [29]. The eigenvectors are the space function part which is composed of several mutually independent and orthogonal space modes [30].…”

Temporal Variation of Sunshine Duration and Its Potential Affecting Factors with Emphasis on Saudi Arabia for the Period 1983-2022

“…In other words, possible spatial modes of variability (patterns of variability) in climate studies can often be studied by EOF analysis. This technique decomposes the time series dataset into spatial and temporal components, to get a better understanding of the variability with minimal modes required [32]. In a space function, the eigenvectors are composed of several orthogonal space modes, each of which is mutually independent [11].…”